Medical Device Link.

IVD Technology Magazine
IVDT Article Index

Originally Published July 2000

Automating POC instrumentation: A systems view

By bringing together advanced componentry and technical sophistication, the current generation of point-of-care instruments is making significant headway toward improving clinical diagnostics.

Among clinicians, the gradual growth of point-of-care (POC) testing represents an opportunity to dramatically improve the quality of patient care for a variety of diseases and conditions. Especially in settings where clinical decisions are time critical, healthcare providers are becoming increasingly reliant on POC devices to help them triage incoming patients, monitor the condition of those patients, and adjust patient treatments.

For manufacturers of such POC instruments, the market potential of the field continues to grow, as clinicians express their desire for ever-more-sophisticated tests. The evolution of bedside glucose testing is a typical example. In the past, only central laboratories were able to provide accurate and reliable glucose-level test results. But when physicians insisted that having real-time glucose measurements was imperative for making clinical decisions, industry responded. With the development of the bedside glucometer, POC glucose testing is now considered the standard of care.

Similar opportunities exist in a variety of fields. In cardiology, for instance, the cardiac markers myoglobin, creatine kinase-MB, and troponin (I and T) are currently used as indicators of a recent acute myocardial infarction. The American College of Cardiology estimates that instituting more-precise protocols for triaging patients—such as the use of POC testing—would save several thousand lives per year, and could also reduce annual U.S. healthcare costs by as much as $6 billion.¹ The cardiac care market is projected to grow at a rate of 40% annually.²

With the unmet needs of such a large market before them, a number of manufacturers are working to develop instrument platforms that can deliver diagnostic test results to healthcare providers at the point of care. To make such devices easy to use, manufacturers are incorporating subsystems that automate a wide variety of functions, including calibration, test operation, and analysis.

One such device now under development is the RAMP system by Response Biomedical Corp. (Burnaby, BC, Canada), which is designed to accommodate many tests that may be needed in near-patient settings. This article looks at how the RAMP instrument integrates a number of complex subsystems to create a highly automated POC analyzer. It is in many ways typical of the current generation of POC devices that are moving into clinical settings at a rapid pace.

A Quantitative POC Platform

The RAMP instrument is a dual-wavelength fluorescence-scanning spectrophotometer (see Figure 1). The system uses inexpensive components to achieve accurate, sensitive information about the quantity and distribution of a fluorescently dyed latex along a nitrocellulose immunoassay strip. The scanning feature enables the system to determine the distribution and quantity of the latex and permits accurate determination of interference and background.

Figure 1. The RAMP instrument by Response Biomedical uses fluorescent spectrophotometry to quantitate analytes on an immunochromatographic test strip contained within a molded plastic cartridge.

In the RAMP system, the immunochromatographic test strip is contained within a molded plastic cartridge. To scan the test strip, the user inserts the cartridge into the reader, which automatically advances the cartridge into position and takes repeated readings of the strip.

Bar Code Reading Made Simple

Before scanning or any other operations can be accomplished, the RAMP instrument must determine whether it has the necessary information about the assay to be performed. This is accomplished by associating a lot code printed on the cartridge with information about that lot stored in the reader's nonvolatile memory. The lot code is provided in the form of a low-resolution five-digit bar code, which is ink-jet printed on the bottom of each cartridge during manufacture. The instrument must read this bar code before it can begin to perform an assay.

The bar code reader is a simple infrared optoreflective detector of the sort typically used as a position sensor, and is available from a variety of sources at low cost (see Figure 2). In the RAMP reader, however, the sensor is modified by masking the optical end of the device to increase its resolution. Although masking causes a loss of sensitivity, there is still more than enough contrast to detect the printed dark bars and the interstitial white signal.

Figure 2. The RAMP system uses an infrared optoreflective bar code reader to identify the test being performed and confirm positioning of the test cartridge.

When a test cartridge is placed into the reader's carriage, as described below, the bar code reader senses its presence. To ensure that the cartridge is in the right position, a position bar is printed at the beginning of the bar code sequence. If the position bar is not in the correct location, the reader reports an error condition and the operator is advised to correctly place the cartridge. The bar code is then scanned using the same stepper-motor positioner that is used for optical scanning. The entire bar code is scanned and decoded.

Cartridge Mechanics

Although the procedure for inserting a test cartridge to be scanned is simple to the user, the mechanical operations behind the scanning system are actually quite complex. When the cartridge is inserted into the reader, it is received into a carriage that is mounted on a linear bushing rod. The carriage is connected to a stepper motor with a precision leadscrew, enabling it to move in a direction parallel to the direction of the sample flow on the test strip. Each step of the motor causes the carriage assembly to move in small increments relative to the optics assembly.

In operation, the reader scans a test strip by taking repeated measurements of the strip in one position, advancing the cartridge incrementally, and then repeating these two steps until the entire binding zone has been measured. This procedure begins when a cartridge is inserted and the carriage is moved to the "home" position. To determine whether the carriage is in the home position, the reader uses an optointerrupt device that senses the position of a flag mounted on the carriage.

As a check on the functioning of this mechanical system, the reader periodically sends the carriage back to the home position. If the optointerrupt device does not sense the flag when the carriage is supposed to be in the home position, the reader reports an error condition. This mechanical check is always performed before the reader reports any test results.

The Optical System

The optical system of the RAMP reader employs a design in which the excitation light path is at a 45° angle to the surface of the target area and the detection light path is at a right angle to the plane of the target. The optical components in both the emission and detection channels consist of an image transfer with an interference filter between the lenses (see Figure 3). The location of the filter allows the light to pass through it at close to a right angle so that there is minimal wavelength-shift due to off-angle incident light.

Figure 3. Optical components of the RAMP system consist of an image transfer with an interference filter between the lenses.

To reduce electrical noise that could interfere with test results, the light-emitting diode (LED) light source and detector are mounted directly on small surface-mount printed circuit boards located within the optical block of the reader. The LED is driven by a capacitor that is located on the LED board so that no high-current signal lines must go to the board. In addition, this board is shielded to eliminate radiated electrical noise. The detector photodiode is mounted on a separate board, and is connected directly to a low-noise operational amplifier using short leads. This practice minimizes the antenna effect of longer leads and results in a low-noise signal.

In order to increase the amount of light produced by the LED light source, the LED is pulsed with a 10% duty cycle. Pulsing the LED allows the current flow during the time the light is on to be significantly greater than the continuous-current rating of the LED. However, pulsing the LED can also result in some signal drift. To compensate for this phenomenon, a separate reference photodiode is used to measure the amount of light emitted by each pulse. Although the amount of light received by this photodiode may vary from device to device, it remains constant for any one device. The reading collected by the reference photodiode is used to correct the driver current during subsequent LED pulses so that output and long-term drift can be minimized.

The Scanning System

To begin the actual scanning operation, the reader advances the carriage to a point before the first peak of the immunochromatographic test line on the test strip. The LED light source is then pulsed on, and the detector receives the resulting signal. These signals are time-gated so that the reader uses only the signal collected when the light is on (see Figure 4). The measured values are converted to a digital format by a 10-bit analog-to-digital convertor and are stored as voltages in a capacitor accumulator. After 50 such readings have been collected, the measured values and information about the position of the carriage are transferred to the microprocessor memory for later processing. The reader then moves the carriage to the next scanning position and clears the capacitor so that the next set of readings can be performed.

Figure 4. In the scanning system of the RAMP reader, signals from the LED light source are time-gated so that the instrument uses only the signal collected when the light is on.

After readings have been completed for the entire binding zone of the immunochromatographic test strip, the reader processes the resulting data. The microprocessor begins by identifying the highest peak value among all the measurements. From this, the reader can calculate the value of the entire "window" of readings. The integrated area of the peak is calculated by simply adding the values of all the stored signals from within the area of the window. The location of the window allows for slight shifts in the position of the test strip without loss of accuracy.

Once the reader has determined the values for each of the two fluorescent labels in both binding zones, it is then able to calculate the final results of the test. To do so, the reader subtracts the value of the red signal at the test line (indicating nonspecific binding in the presence of the capture antibody) from the value of the blue signal at the test line (indicating actual binding of the analyte antibody). This calculation corrects for nonspecific binding on a test-by-test basis, and is a feature unique to the RAMP system. This calculated test-line value is then divided by the total value of the red and blue signals at the internal standard line to correct for such assay variations as membrane thickness, sample viscosity, and reaction speed. This test-by-test correction is the key to providing accurate quantitative results from a lateral-flow assay, and is also a key element of the RAMP patent.

In addition to calculating the positions of the peak values, the reader also determines the positions of two locations where the signal should be at background levels. If the measurements at those two positions are too high as a result of high background binding in the assay, the reader reports an error condition. If those background measurements are low, the values are averaged and then subtracted from the reading at the peaks, thus correcting the actual signal for nonspecific background binding.

The RAMP reader incorporates a number of features that contribute to the accuracy and reliability of each test. For instance, the time-gated signal collected while the LED is off can be used to help determine the optical and electrical background of the test, and thereby to identify and correct conditions that are outside specified test parameters. The reader's ability to report conditions in which a test's background is too high, or to correct for nonspecific binding in an individual sample, are also unique to the RAMP format.

Calibration Information

To ensure proper calibration of the RAMP system for each test, the reader is designed to automatically compare the bar code on each test cartridge to lot information stored in the instrument's memory. If the lot is a new one, the instrument prompts the operator to insert the proper information for that lot.

The lot information is stored on and entered by use of a proprietary calibration device called the Cal Stick, which is a simple electrically programmable read-only memory (EPROM) embedded on a small printed circuit board (see Figure 5). The circuit board has electrical contacts which connect to the instrument when it is inserted. The Cal Stick contains information about the lots of assay cartridges, including the name of the assay, its lot number, the expiration date of the lot, the characteristics of the calibration curve, signal versus analyte concentration, and measurement units to use in reporting results (e.g., ng/ml).

Figure 5. Calibration information for the RAMP system is entered by means of an EPROM embedded on a small printed circuit board.

Once the Cal Stick has been inserted, the information is downloaded and stored in the reader's nonvolatile memory. The Cal Stick only needs to be inserted once. The reader's memory can store information about as many as 50 lots; if more than 50 lots are used, the oldest one will be erased and overwritten. If the inserted Cal Stick provides information about a lot number that is the same as an existing record, the older information will be overwritten. This option can be used to provide updated information about a lot that is already in the field.

User Interfaces

In the current generation of POC devices, automated functions often extend well beyond instrument subsystems and into software-driven features. Such features offer operational enhancements that are especially important for POC devices, which are often designed for use by healthcare professionals without significant clinical laboratory training or experience. Creating instruments that are easy to use, even in untrained hands, often requires considerable design and engineering sophistication, but the result can be a startling improvement over earlier products.

The RAMP system includes a number of such features that go beyond the basic tasks of scanning the test cartridges and reporting the assay results. One useful feature, for instance, helps to protect operators from the distractions of their environment by prompting to complete a test once it has been initiated. As part of the normal operation of the instrument, after the patient identification has been entered, the reader prompts the operator to add sample to the RAMP cartridge and place the cartridge into the reader. If a cartridge is not inserted within 5 minutes of entering the patient ID, an alarm sounds and a display message again advises the operator to place the cartridge into the instrument. The instrument repeats this prompt for another 5 minutes, after which it resets itself to the main menu and removes the patient ID from memory.

Another feature helps to ensure that operators supply enough patient sample for the reader to perform each test. When a cartridge is inserted, the instrument reads and validates the bar code, as described previously. Then the reader moves the carriage to a position at the end of the assay strip to await completion of the wicking step of the immunochromatographic assay. Although the wicking action normally takes about two minutes, it can take longer if an insufficient amount of sample is added to the test cartridge. In the RAMP system, if the liquid front has not reached the end of the test strip after four minutes, the reader reports an error condition and informs the user that insufficient sample was added.

The RAMP system also provides operators with current information about the status of a test in progress. After the wicking portion of the reaction is complete, a "development" time is needed before the scan can be performed. The reader thus starts a 6½-minute timer, which is displayed as a countdown on the instrument's screen. When the development cycle is finished, the strip is scanned as described earlier. During the scan, the reader reports any errors that involve the background levels of the test or the position of the test cartridge.

The reader displays assay results as soon as it has completed verification of the scanning and positioning data. The displayed result consists of the patient ID, the test name and result, units of measure, the date and time the test was performed, the lot number and expiration date of the assay kit used, and (optionally) the operator ID. The reader also stores this test result information in its nonvolatile memory, which can accommodate up to 100 sets of results.

Finally, the RAMP reader can also be connected to a laboratory information system (LIS) for uploading test results. The main menu permits the user to upload some or all of the 100 results stored in memory. This is accomplished by connecting the reader to a standard RS-232C serial port. On command, the reader sends a data stream to the LIS. There is a handshake verification of the communication. If desired, a macro in the LIS can file the information by patient ID, test type, days or dates the tests were run, or operator ID.

Conclusion

Many types of clinical testing offer opportunities for continued growth of the POC market. Wherever healthcare personnel must triage, monitor, and treat patients on the basis of time-sensitive results, POC devices are a natural choice.

Manufacturers continue to improve the operation of such instruments, with the goals of making them faster, more specific, more sensitive, and easier to use. Such improvements will continue to rely on careful attention to the integration of device subsystems such as those described in this article.

References

1. Ryan et al, "Management of Acute Myocardial Infarction," Journal of the American College of Cardiology 28, no. 5 (1996):1328–1428.

2. Gibbons, W, Equity Research Report (Chicago: William Blair & Associates, 1998).

3. Cardiac Markers, ed. Alan H. Wu (Totowa, NJ: Humana, 1998).

Photos Courtesy Response Biomedical

Paul Harris is vice president for R&D and Joanne Stephenson is vice president for business development at Response Biomedical Corp. (Burnaby, BC, Canada).